[Info-vax] STARTREK.BCK Restore - How?

Stephen Hoffman seaohveh at hoffmanlabs.invalid
Fri Jun 14 18:31:11 EDT 2019


On 2019-06-14 15:38:35 +0000, Tom Wade said:

> On 2019-06-12 15:56, Stephen Hoffman wrote:
> 
> 
>> Yet I've encountered BACKUP savesets of those VMSINSTAL kits, and the 
>> folks mess up that omnibus BACKUP during the transfer.
> 
> If you mean taking the VMSINSTAL backup sets (.A, .B etc) and putting 
> it into another .BCK and subsequently ZIPping, then yes, I agree this 
> is silly.  My preferred way to ship VMSINSTAL kits is to use Backup to 
> create the .A, .B etc then put them all into a single ZIP/VMS along 
> with a COVER.TXT.  You did, say, however, you saw no need for Backup to 
> be involved at all.

If this is the STARTREK.BCK saveset from the title, I'm assuming y'all 
are using VMSINSTAL here because PCSI is privileged, and because 
OpenVMS really doesn't have a good non-privileged software installation 
mechanism.

As for PCSI and VMSINSTAL, they're both weak sauce.

And I'd ponder whether there are still enough folks that are working 
with OpenVMS systems and that are loading games that don't also have 
system privileges.

> Now if only Backup could write compressed savesets as 512 byte fixed 
> records (/DATA=INTERNET ?).

Preferably, BACKUP needs to deal with whatever saveset it got handed. 
Not more qualifiers to special-case saveset creation.  Nor extra 
qualifiers to repair the saveset.  No more documenting around the error 
and the limits.  BACKUP gets handed a file, and it deals with whatever 
it got handed.

OpenVMS itself and many OpenVMS apps just haven't been good at that 
sort of design for a while, though.

BACKUP has a baroque user interface, a design that makes fixing this 
and related issues far more difficult, and the current BACKUP I/O 
design is at diminishing returns; out of headroom.  As I've commented 
previously.

Writing code for OpenVMS is comparatively more difficult and with more 
dependencies, whether it's CLI parsing, or it's preferences and 
configuration files, or LDAP integration, or the development tooling, 
or whatever.

The entirety of "software distribution" and patch management on OpenVMS 
is somewhere between utterly hilarious and just plain sad, and the 
related tools—VMSINSTAL and PCSI and DCL itself—are weak, at best.

VSI has a whole pile of work ahead here, once the port is out the door 
and available.

But again, I see no reason for BACKUP to be involved here as a software 
distribution mechanism for STARTREK-the-game, or whatever.  Yes, you're 
somehow stuck with it for VMSINSTAL, but then that tool really needs to 
follow VMSUPDATE into oblivion.


-- 
Pure Personal Opinion | HoffmanLabs LLC 




More information about the Info-vax mailing list